US3649350A - Electroless copper plating - Google Patents

Abstract

Known electroless copper plating solutions comprising basic aqueous solutions of a cupric ion source, a complexing agent for cupric ion and a reducing agent for reducing cupric ion to copper metal, which are normally stabilized by aeration, are further stabilized by incorporation of certain water-soluble salts of hydroxy aliphatic hydrocarbon sulfonic acids. An improved quality of copper deposit is obtained from these stabilized baths.

Description

United States Patent Agens [54] ELECTROLESS COPPER PLATING [72] Inventor: Maynard C. Agens, Burnt Hills, NY. [73] Assignee: General Electric Company [22] Filed: June 29, 1970 [21] Appl. No.: 50,996

[52] U.S.C1 ..l17/l60 R, 106/1, 117/47 A,

117/47 R, 117/130 E,117/212,117/213,117/227 [51] Int. Cl v.C23c 3/02 [58} Field ofSearch ..117/130E, 160R; 106/1 [56] References Cited UNITED STATES PATENTS 2,903,403 9/1959 Strauss ..l06/l X 3,222,195 12/1965 Pearlstein ..106/1 3,361,580 l/l968 Schneble, Jr. et al. ....l 17/130 E X 3,415,666 12/1968 Nagaiet al ....117/130EX 3,420,680 l/1969 Gulla v...1 17/130 E X 3,436,233 4/1969 Jackson ....117/130 E X 3,453,123 7/1969 Merker et al. ...117/13OEX Mar. 14, 1972 3,454,416 7/1969 Heymann et al ..1 17/130 E X 3,457,089 7/1969 Shipley, Jr. et al. ..l17/130 E X FOREIGN PATENTS OR APPLICATIONS 1,184,123 3/1970 Great Britain 106/1 Primary ExaminerAlfred L. Leavitt Assistant ExaminerJ. R. Batten, Jr.

Attorney-James W. Underwood, Richard R. Brainard, Paul A. Frank, Joseph T. Cohen, Frank L. Neuhauser, Oscar B. Waddell and Joseph B. Forman 5 7] ABSTRACT Known electroless copper plating solutions comprising basic aqueous solutions of a cupric ion source, a complexing agent for cupric ion and a reducing agent for reducing cupric ion to copper metal, which are normally stabilized by aeration, are further stabilized by incorporation of certain water-soluble salts of hydroxy aliphatic hydrocarbon sulfonic acids. An improved quality of copper deposit is obtained from these stabilized baths.

10 Claims, N0 Drawings 1 ELECTROLESS COPPER PLATING The present invention relates to improved, stabilized, electroless copper plating solutions and to a process for using these solutions for the copper plating of substrates having a surface which is catalytically active in causing deposition of copper from electroless copper plating solutions.

The present invention has for its object the provision of improved electroless plating baths which are simple to use, extremely stable, are highly reliable, economical to use and can be used for long periods of time to deposit any desired thickness of copper and the chemicals, which are depleted by the plating reaction, are readily replenished to pennit extended use of the solutions.

A further object of the invention is the provision ofa novel and improved process for the electroless plating or deposition of an improved quality of copper on various substrates, for example insulating substrates for electrical circuits, metal, ceramic or other desired supporting surfaces. The improved quality is generally manifested in the physical properties of the copper layer, for example its ductility, color, etc.

A wide variety of electroless copper plating baths and processes have been described in the prior art and many of them have been placed in commercial operation for the deposition of copper on various substrates for either decorative or utilitarian purposes. Electroless copper plating baths, by their very nature must be somewhat unstable in order to copper plate on the catalytically active surface. Many of the earlier solutions were very unstable and after a relatively short useful life, decomposed depositing copper indiscriminately. In an attempt to overcome this catastrophic decomposition during use, various techniques have been developed to increase the stability of the bath, for example, by use of very strong complexing or chelating agents for cupric and cuprous ion, incorporation ofvarious additives, etc.

While such alternatives increase the stability of the bath against autogeneous decomposition, they have the undesirable effect that they greatly increase the length of time required to deposit a given thickness of copper plate because the stability of the bath is generally directly related to the rate at which copper is deposited from the bath. For this reason, electroless plating of copper is used in many applications to deposit only very thin copper films to provide an electrically conductive surface which is built up to the desired thickness by electrodeposition of copper. Where an entire surface of a substrate is being plated, this generally causes no difficulties. However, when it is desired to deposit relatively thick layers of copper on a multitude of isolated metal areas, for example, to make a multiple circuit board on an insulating substrate where there can be hundreds of isolated areas of copper, the connection of an electrical lead to each of these areas to permit buildup of the copper deposit by electrodeposition on all of these areas presents an almost insurmountable and, at best, a very expensive task. It would be highly desirable to be able to plate the entire circuit to its desired thickness using electroless plating techniques, thereby, eliminating the need for the electrical connections.

One method, which has found wide use in industry for the stabilization of electroless copper plating solutions, has been the aeration of the electroless plating solution using air or other oxygen containing gas. This approach to stabilization overcomes many of the disadvantages of other means of stabilization of electroless plating solutions in that it does not have any adverse effect on the rate at which copper is plated from the stabilized solutions. This process of stabilization is fully described and claimed in my U.S. Pat. No. 2,938,805 which is hereby incorporated by reference.

The rate of plating of copper from the electroless copper plating solutions is also dependent on the actual copper ion concentration in the solution. This means that after the copper ion concentration has been partially depleted through use, the rate becomes so slow that it is no longer economical to continue to use the solution although even at this point there is considerable copper ion still remaining in solution. In the past,

it has been usual to discard the solutions at this point and start with fresh solution. Attempts to replenish the copper ion concentration and other chemical components depleted by the plating reaction up to their initial levels, has been attempted. Generally the stability of these reconstituted solutions is poorer than the initial solutions so that regeneration is only marginally economical.

A few electroless copper plating solutions are sufficiently stable that it is economical to replenish the used chemicals. However, the chelating agents or additives used to produce the required stability, are so effective that plating on the catalytically active surfaces does not occur at ambient temperature and heating of the solution is required to induce plating. Even then, extremely long periods of time are required, in the order of 24 hours or more, to electroless plate a one mil thick coating of copper.

I have now discovered that extremely stable electroless copper plating baths can be made by adding certain hydroxy sulfonates to any of the previously known electroless plating solutions comprising a source of cupric ions, a source of hydroxide ions sufficient to provide the pH required for electroless copper plating, generally, a pH greater than l0 and preferably l2 to 13, an agent capable of reducing copper ions to copper metal in alkaline solutions, generally, formaldehyde or a formaldehyde engendering material, e.g., paraform, etc., in the presence of a surface which is catalytically active in causing deposition of copper from an electroless copper plating solution, generally, either a metal surface itself or a non metallic surface which has been rendered catalytically active by deposition of metallic nuclei either from a colloidal solution of metal or by absorption of metal nuclei by first treating with a reducing solution for example an acidified aqueous solution of stannous chloride followed by rinsing and treatment with an acidified aqueous solution of a noble metal salt, for example, palladium or platinum chloride, etc.

The electroless copper plating solutions which I may use are well known in the art. They are described in both the patent and technical literature. Typical, but not limiting. of this art are US. Pats. No. 2,874,072; No. 2,938,805; No. 2,996,408; No. 3,075,855; No. 3,075,856; Agens and Kantor copending application, Ser. No. 8l 1,012, filed Mar. 27, 1969 and assigned to the same assignee as the present invention, and the various literature and prior art cited and referred to in these patents. The teaching of all of this art with regards to their compositions and techniques of use are hereby incorporated by reference.

This above-described stability is accomplished by adding to the electroless copper plating solution a water-soluble salt selected from the group consisting of dithionites, bisullite adducts of lower alkyl aldehydes and bisulfite adducts of unsaturated lower aliphatic hydrocarbonols. By the term unsaturated lower aliphatic hydrocarbonols, I mean hydroxy substituted hydrocarbons of the aliphatic series having from one to eight carbon atoms and having either olefinic or acetylenic unsaturation. These compounds are olefinic and acetylenic alcohols, including polyhydric alcohols, of the aliphatic series.

As chemical compounds, these materials and their method of preparation are well known in the art. Dithionites, also known as hyposulfites or hydrosulfites, generally made as the sodium salt, are used in dyeing. They react with formaldehyde to form an equimolar mixture of the corresponding hydroxymethane sulfinate salt and hydroxymethane sulfonate salt. The latter compound is the same compound as the bisulfite addition compound of formaldehyde.

The addition of bisulfites to ketones and aldehydes is an equilibrium reaction which is shifted by pH. A high pH, causes a shift to the free aldehyde or ketone and the bisulfite and is much more pronounced for ketones than aldehydes. When a bisulfite adds to an olefinic double bond it forms a saturated sulfonate and when a bisulfite adds to an acetylenic bond it can form an olefinic sulfonate or a saturated disulfonate depending on whether one or two molecules of bisulfite add to the triple bond. The reaction products in both cases are probably a mixture of the possible stereoisomers since the sul fonate group can add to either one of the two unsaturated carbon atoms. These adducts to olefms and acetylenes are not in equilibrium as are the adducts of ketones and aldehydes.

Because of the above-described equilibrium reaction, and because of the high pH of electroless copper plating solutions, the bisulfite adducts of ketones are oflittle, if any value, as additives to electroless copper plating solutions and the aldehyde adducts, although satisfactory are not nearly as effective as the bisulfite adducts of olefinic and acetylenic hydrocarbon alcohols for my purpose. The exact function of the hydroxyl group is not known but it is necessary. It can be supplied by a hydroxyl engendering group for example, an ester group which hydrolyzes under the basic conditions of the bath to produce the hydroxyl group on the moiety also having the sulfonate group. Polyhydric alcohols are better than monohydric alcohols. There may be more than one olefinic or acetylenic group in the alcohol molecule from which the bisulfite adduct is made.

Since the hydrocarbon moiety of the hydroxy sulfonates does not have any effect there is no purpose in using any hydroxy sulfonate where the hydrocarbon moiety has more than eight carbon atoms, i.e., the moiety should be a lower aliphatic hydrocarbon. Likewise, the metal cation does not have any effect on the efficiency of these hydroxy sulfonates for my purpose as long as they do not make them insoluble in the plating solution. They can be alkali metal or alkaline earth metal cations, but are preferably the former, especially sodium because ofits cheapness and availability.

Typical, but not exhaustive, of the art showing, various bisulfite adducts useful in my invention are US. Pats. No. 2,793,229; No. 2,820,818; No. 2,857,427; No. 3,211,783; British Pats. No. 756,105; No. 774,563; the article by Reppe et al. in J. Liebigs Ann. 596, l (1955) etc., and the references cited in this art. These teachings are hereby incorporated by reference. For some unknown reason, the bisulfite adducts obtained by filtration from the reaction mixture are apparently more effective for my purpose than the same material which has been further purified by recrystallization.

Only a very small amount of the water-soluble salt of the above materials is required to effectively stabilize the electroless copper plating solutions against autogeneous decomposition, generally amounts as low as 0.005 g./l. of plating solution will produce a noticeable stabilizing effect with best results obtained in the range of 0.01 g./l. to 0.08 g./l. The optimum amount is in the range of about 0.02 to 0.04 g./l. Even higher amounts can be used but there is no incentive to do so since the plating rate is noticeably decreased as the amount of my additive is increased and the quality of the copper deposit decreases. However, amounts as high as l g./l. do not stop the plating solution.

It was indeed surprising to find that these materials were ef fective stabilizers in such extremely small amounts, but it was even more surprising to find that with this increased stability, the plating rate of the stabilized solutions was affected little if at all. The solutions readily plate copper on activated surfaces at room temperature, will stop plating when the activated surface is removed from the solution and will not decompose on standing for long periods of time, especially if aerated. The depleted chemicals can be readily replaced and additional plating performed for a period of many days with no evidence of the bath autogeneously decomposing. It was still more surprising to find that the quality of the copper deposit was also beneficially affected, perhaps due to the controlled deposition of the copper from the electroless plating bath. The plated copper has a very bright color and a very fine grain texture. When a very smooth metal surface, for example a stainless steel sheet, or a plastic laminate which has been only lightly vapor blasted, is electroless copper plated, which permits nondestructive stripping of the copper plate from the substrate, the copper is found to be extremely flexible, much like that of electrolytically produced copper sheet. When my elcctroless copper solution is used to produce through-hole plating and electrical circuits on an insulating substrate and the various electronic components mounted in the holes and soldered in place by passing the circuit board over the surface ofa molten bath of solder in a wave soldering machine, the components are readily soldered to the circuit board with no failures in the electrical circuit. Furthermore, the circuit components can be unsoldered and new components soldered in place without failure or damage to the electrical circuit.

In order that those skilled in the art may better understand my invention, the following examples are given by way ofillustration and not by way of limitation. Unless stated otherwise, one liter of plating solution was used to copper plate both sides of a 1.5"X4" epoxy resin bonded glass laminate plaque that was activated by the standard stannous chloride-pailadium chloride technique. When used, the stainless steei plaque was also the same size and was similarily activated.

STANDARD SOLUTION The solution chosen for evaluating the effect of my additives which is typical of the electroless copper plating solutions of the prior art is as follows:

CuSO, H 0 25 gJl. Nichol-I 0 l g./l. Rochelle SaItAH O 37.5 g./l. NaOH l6 gJl. Formaldehyde (37%) i5 mL/l.

This solution will readily plate copper onto metal surfaces or activated nonmetallic surfaces. Generally the formaldehyde is added just prior to use. However, as is typical of the prior art electroless copper plating solutions, this solution will autogeneously decompose, precipitating copper throughout the solution and onto the walls of the container after several hours of use and must therefore be discarded at that time. The copper plate obtained from this solution has a good appearance but, if stripped from a high mirror substrate such as stainless steel, it will crack if severely bent and creased. Aeration will definitely increase the life of this plating solution but even then the solution will decompose on standing overnight at ambient temperature. Aeration with a stream of very fine bubbles of air was used in all examples except where stated.

EXAMPLE 1 Using the above standard solution, various hydroxy sulfonates were added and epoxy bonded glass laminate samples which had been only lightly vapor blasted prior to activation by the standard stannous chloride-palladium chloride treatment were introduced into each one of the solutions and allowed to plate overnight by which time an approximately 1.5 mil coating of copper had plated on the surface of the laminate. The rate of plating decreases with time because of the decrease in copper ion concentration as plating proceeds. Generally, a one mil coating of copper was obtained in the first five to seven hours. A fresh piece of the laminate was introduced and the plating reaction continued for a cumulative total plating time of 24 hours to evaluate the stability. My work with these hydroxy sulfonates has shown that if no autogeneous decomposition has occurred in 24 hours, none would occur no matter how long plating is continued.

A one inch wide strip of copper was removed from the glass epoxy laminate plated overnight and subjected to a bend test. ln this bend test, the strip of copper is formed into a loop which is placed under a weighted roller running in a one inch wide track. Passing of the roller over the strip, forms a crease. The strip was then opened and the crease flattened by again running the roiler over the strip. These two operations constituted one bend in the measurement and were repeated until the copper separated at the crease. Designation of the test results as a half bend indicates that no failure occurred during the crease formation but failure did occur during the second portion of the test. For example, in the following table, a result of L5 in the bend test indicates that the strip was creased once the crease flattened once, the time, all without failure with crease reformed the second failure occurring in the second was chosen to no any decomposition of the plastic substrate nor oxidation of the Copper.

Table 1 shows results of 24 hour period and had copper particles precipitated in the solution and would no longer plate copper onto a catalytic surface. They are typical of a solution which had undergone catastrophic decomposition after a period of use.

Sodium g./l. added Bend test Sodium hisulfite to plating not suit adduct of solution Stability annealed annealed 2-propane sulfonate 04 U l-hutune sulfonate 0.04 U Z-hydroxy Z-propane acetone 0.04 U stlllbnltlt: l-hydroxyproprion' 0.04 P.S. l 2 15-2 l'prupunc aldehyde 0.08 PS. 3 Z.5-3 sulfttnale 3-hydroxytillyl 0.04 P.S. l5 l5 l-propane alcohol 0.08 S 2.5 2.5 sullottute Lhydroxy prop-urgyl 0.04 S 1.5 1.5 l.2 alcohol 0 005 5 L5 1.5 propane disulfonate 1.4dihy- Z-butenc 0.04 S 2.5 3.5 droxy-2 lshdiol 0 S 8 butane sull'onate l 4-Liihyl,4 butyne- 0.04 s 3 4 droxy2,3- diol 0.02 S 3 4.5 butane (diztdduct) disull'nnute l 4-dihyl,4-hutynedroxy l diol 0.02 S 1.5 L5 hut-Zene tmono' sulfonatc adduct) results. The first two entries lack any hydroxyl group showing the necessity for the presence of a hydroxyl group in the sulfonate molecule. The third entry illustrates that the bisulfite adducts of ketones as compared to aldehydes are likewise not satisfactory. Of all the ketones, acetone forms the most stable bisulfite adduct.

EXAMPLE 2 ln Example 1, the bisulfite adducts were isolated as white crystalline compounds prior to use but were not purified by recrystallization. In this example, the bisulfite adduct is preformed but not isolated from solution because sufficient water is used to keep it in solution. A solution of 20.8 g. of sodium bisulfite, 8.6 g. of l,4-butynediol in 200 cc. of water was heated to boiling and then cooled to room temperature. No crystals formed on cooling. A 10 ml. aliquot of this solution was added to one liter of the standard plating solution which was slightly modified by increasing the amount of formaldebyde to 20 g./l. This increase in formaldehyde has no of feet on the stability or quality ofthe copper plating but merely permits more copper to be plated. An epoxy bonded glass laminate and a stainless steel sheet were placed in this bath and plating allowed to continue overnight. The bath remained completely stable and the copper plate when removed from the stainless steel and submitted to the bend test, was capable of taking 2.5 bends before annealing and five bends after annealing.

When a 10 ml. aliquot of a solution of 8.6 g. of 1,4-butynediol in 200 ml. of water was used in place of the i0 ml. aliquot of the bisulfite adduct to this diol, in the above solution, the plating bath catastrophically decomposed after two hours 01 plating. When this same test was repeated using a l0 ml. aliquot of a solution of 20.8 g. of sodium bisulfite in 200 ml. of water, the bath did not catastrophically decompose but did permit some plating of the copper on the container. It should be recognized that the sodium bisulfite can react with the formaldehyde of the solution to form the sodium bisulfite adduct of formaldehyde and this would account for this stability. However, the highly alkaline conditions of the electroless copper plating bath are not the ideal conditions for making the sodium bisulfite adduct of formaldehyde. As previously mentioned, dithionites react with formaldehyde giving two products, one of which is the same as the bisulfite adduct of formaldehyde. The following example illustrates the use of this material.

EXAMPLE 3 When an attempt is made to dissolve sodium dithionite in water, the solution immediately becomes cloudy but when dissolved in a formaldehyde solution, the solution remains clear. A solution of two g. of sodium dithionite in 1 liter of 37 percent aqueous formaldehyde solution was prepared. A 15 ml. aliquot of this solution was used as a replacement for the 15 ml. of formaldehyde in the standard plating solution. The standard laminate was plated in this solution overnight. copper strip was capable of taking two bends in the bend test and the bath was completely stable and showed no evidence of decomposition or extraneous plating after a 24 hour period. used and solid sodium mg./l. likewise produced extremely stable solutions. The copper when stripped from the laminate was capable of 1.5 ybends before annealing and 6.5 bends after annealing.

EXAMPLE 4 This example illustrates the effect of variation of the amount of the adduct and the ability of the solution to be used without aeration for an extended period of time. The adduct was the sodium bisulfite diadduct of 1,4-butynediol, dissolved in water to give a concentration of four g./l. One liter aliquots of the standard plating bath were used in which various amounts of the solution of the adduct were added. The results are shown in Table II.

EXAMPLE This example illustrates the extended lifetime of the copper plating solutions and the ability to replace the depleted chemicals to further extend the useful lifetime of these plating solutions. One liter of the standard plating solution was used to which was added a ml. aliquot of the bisulfite diadduct of 1,4-butynediol made up to contain four g./l. The standard laminate was used and allowed to plate for 23.75 hours. One side of the laminate had a 1.67 mil coat of copper and the other side had a 2.04 mil coat of copper. When stripped, the copper was capable of standing 3.5 bends without annealing and 4.5 bends after annealing.

The chemicals which had been depleted from the bath were replaced by using replenishing solutions A and B, whose compositions were as follows:

tetruhydrate HCHOUVR) l64 rnl/l.

The composition of these two replenishing solutions are based on calculations of the amount of chemicals used in plating a given amount of copper. Copper can readily be determined by a colorimeter to determine the amount of copper that has to be replaced in order that the volume would not continue to grow larger, a volume of the depleted plating bath equal to the volume of the replenishing chemicals is removed prior to adding solutions A and B. The concentrations of A and B are such that 2.5 volumes ofA are used per volume of B.

After the above plating, the solution was replenished by first removing 95.8 ml. and adding 68.4 ml. of A, 27.4 ml. ofB and five ml. of the adduct solution. The solution was again allowed to plate copper on the standard laminate for 23.75 hours. The thickness of the copper on one side was 1.56 mils and on the other side 1.66 ml. When stripped from the laminate, the copper was capable of taking 1.5 bends before annealing 7.5 bends after annealing for 20 minutes at 120 C. The solution was again replenished by removing 95.8 ml. of the spent plat' ing solution and replacing it with 68.4 mi. of solution A and 27.4 ml. of solution B and adding a 10 ml. aliquot of the ad duct solution. The standard epoxy giass laminate was added and plating allowed to continue for 7.5 hours. The thickness of the copper plate was one mil on one side and 1.02 mil on the other side. This copper when removed from the laminate was capable of taking 55 bends both before and after annealing.

Again the solution was replenished by removing 92 ml. of the solution and replacing it with 65.7 ml. of solution A and 26.3 ml. of solution B and adding an additional 10 ml. aliquot of the adduct solution.

After plating, a standard glass laminate for l7.25 hours, the solution was again replenished by removing 100.9 ml. of the solution and replacing it with 78.5 ml. of solution A and 3l.4 ml. of solution B and adding an additional 10 ml. aliquot of the adduct solution. A standard laminate was plated for 24 hours. The solution was replenished again by removing 1 16.6 ml. of the solution and replacing it with 83.3 ml. of solution A and 33.3 ml. ofsolution and adding an additional 10 ml. of the ad duct solution. At the end of an additional 24 hours of plating on a standard laminate panel, the solution still showed no evidence of decomposition or extraneous plating and was still plating a very high quality copper on the glass laminate.

EXAMPLE 6 This illustrates that the rate of plating is faster initially and decreases as the chemicals become depleted in the solution. It also illustrates that aeration is not specifically required although it is preferably used. The standard solution described above was used to which was added 0.04 g./l. of the sodium bisulfite diadduct of 1,4-butynediol. The standard activated laminate was placed in one liter of the solution without aeration for a period of five hours. The copper plated laminate was removed and found to have [.03 mils of bright, flexible copper plating on its surface. A second seeded standard laminate was placed in the bath and plating allowed to continue overnight, still without aeration. The following day, the bath was still stable with no evidence of decomposition. Although the amount of copper which had plated overnight produced only a thin film of copper on the laminate, it was very bright and very flexible.

One of the critical requirements of electroless copper plating solutions is the pH. ifa solution is made up with all the essential ingredients except the alkali, generally sodium hydroxide, which is added in only an amount to give a pH of 9.5, the solution will not plate copper. The pH must be increased to at least 10 before plating will be initiated and at least 12 before a reasonably fast rate of plating is attained. If the pH is increased above about l3, the stability of an otherwise stable solution is affected.

Du ring plating, one of the products of the chemical reaction must be acidic in nature, since monitoring a solution during plating will show a decrease in pH, thus slowing the speed of plating. Generally, adding more alkali to raise the pH will cause a noticeable increase in plating rate over the rate measuredjust prior to the pH adjustment.

It would be desirable to be able to increase the amount of alkali that can be tolerated in the bath without adversely affecting the stability. l have found that this can be done by adding a salt, which with the alkali present in the bath, produces a buffering action at the high pH present in these baths. Any of the well known buffers for a pH greater than 10 can be used. A particularly good salt to add is sodium boratev Boric acid or other metal borates can be used, but in the presence of the large quantity of alkali present in these baths, generally sodium hydroxide, these materials would be converted to or act the same as sodium borate in the solution.

This example shows the use of sodium borate to permit higher concentration ofalkali to be used.

EXAMPLE 7 Two one liter copper plating solutions were prepared using the standard solution except the amount of sodium hydroxide was increased to 25 g./l. Both solutions contained 0.04 g./l. of the sodium bisulfite diadduct of 1,4-butynediol. In addition, one solution contained 38.l g/l. ofsodium borate. A standard activated laminate was inserted into each of the solutions. After five hours of plating with aeration, the average thickness of the laminate plated in the bath without the borate was one mil of copper and 0.7 mil from the solution containing borate. Both solutions were completely stable and had not deposited any extraneous copper. The copper which had plated from the bath containing the borate was brighter and more flexible than that from the solution without borate. Another set of standard laminates seeded for electroless plating were inserted in the baths and allowed to plate for an additional three hours with aeration during which time there was no evidence of instability of either bath. The plating rate had decreased so that the average thickness of the copper plating on the laminates from both baths was 0.26 mil showing that the decrease in plating rate with time had not been as great in the solution containing the borate buffer as had been in the solution not containing the borate buffer.

Although the above examples have all used Rochelle salt as the complexing agent for cupric ion, 1 have successfully stabil ized other electroless copper plating baths containing other complexing agents for the cupric ion, for example, the electroless copper plating solutions disclosed in the above-referenced patents. However, l have noted that when the complexing agent contains nitrogen, for example, ethylenediaminetetracetic acid or its alkali metal salts, ethanol amines, nitriloacetic acids, etc., the copper is not as flexible as when I use a nonnitrogen containing complexing agent. Of all the complexing agents I have tried, I have obtained the optimum results with my stabilizing agents when the complexing agent for the cupric ion has been Rochelle salt. I have also found that the inclusion ofa small amounts ofa nickel salt aids in increasing the quality of the copper, especially the ductility. Although nickel alone can not be plated from these baths when all ofthe copper salt is replaced by a nickel salt, nickel is found by analysis in small amounts in the copper plated from the solution of the specific examples which contain nickel. The above specific examples therefore represent the best mode of practicing my invention.

To illustrate the use of my solutions for making a circuit board containing plated holes, the following example is given.

EXAMPLE 8 A glass fabric laminate having copper foil bonded on each of the two major surfaces, whose individual laminae were bonded together with an epoxy resin containing two percent by weight of stannous oxide was used. This laminate and its use in the making ofelectrical circuits is disclosed and claimed in my copendingjoint application with S. W. Kantor, Ser. No. 811,012, previously mentioned. A multiplicity of electrical circuits were etched on both surfaces using a standard photosensitive process to mask the desired areas and standard etching techniques to produce the desired pattern. After coating the entire surface of the laminate with a protective coating, holes were drilled through the laminate at the desired places in the circuit where it was desired to insert circuit components or to make connections with the circuit on the opposite face of the laminate. After blowing the dust from the holes with air, the laminate was immersed for one minute in a 0.1 percent aqueous solution of palladium chloride containing about 10 g./l. of aqueous hydrochloric acid, heated to l40 F. Only the walls of the drilled holes were activated by this treatment since only in these places was the stannous oxide exposed so that it would reduce the palladium chloride to palladium metal. After rinsing, the palladium which had deposited on the edges of the copper foil exposed in the walls of the hole was removed by immersing the laminate in an aqueous solution of 240 g./l. ammonium persulfate heated to 90 F. After rinsing in distilled water, the laminate was placed in a two-liter bath whose composition was the same as a stan dard solution containing 0.04 g./l. ofthe sodium bisulfite diadduct of 1,4-butynediol. The bath was aerated and plating continued for hours, which produced a very bright and thick deposit of copper in all of the holes which had been drilled. The masking material was stripped from the board and the copper cleaned by immersing in the above ammonium persulfate solution for 30 seconds then rinsing and drying, after which the entire circuit was tinned using a conventional immersion tin plating solution. Circuit components are then readily placed in a desired position in the circuit pattern and soldered in place using a standard wave-soldering machine. When faulty components are detected either during testing of the circuit or during use, such components are readily unsoldered from the circuit board and replaced with a satisfactory component which can be readily soldered in place without any danger of disrupting the electrical circuit of the electroless deposited copper.

Although the above examples have illustrated many modifications that can be made in this invention, other variations will be readily apparent to those skilled in the art. My work has shown that the Rochelle salt, used in the above examples as the complexing agent for cupric ion, can be replaced with any of the other wide variety of chelating agents for cupric ion known in the art and previously used in electroless copper plating solutions. The hydroxy sulfonate compounds ofthis in vention are equally applicable to stabilizing such solutions. The copper plating can be over the entire surface of the laminate or only on selected ortions thereof to roduce either decorative or utilitarian esigns for example 0 ectrical circuits. In addition to plating only on the surface of the laminate, holes may be drilled therein and the walls readily plated where it is desired to make connections between two copper areas on the two major surfaces of the laminate. These and other variations will be readily apparent to those skilled in the art and are within the full intended scope of the invention as defined by the appended claims.

What I claim as new and desire to secure by Letters Patent in the United States Patent Office is:

1. In an electroless copper plating solution comprising a source of cupric ion, a source of hydroxide ions sufficient to provide the pH required for electroless copper plating, an agent capable of reducing copper ions to copper metal in the presence of a surface which is catalytically active in causing deposition of copper from said solution and a sufficient amount of complexing agent to prevent precipitation of cupric hydroxide from the solution, the improvement comprising the addition ofa watersoluble salt selected from the class consisting of dithionites, and bisulfite adducts of lower alkyl aldehydes and bisulfite adducts of unsaturated lower aliphatic hydrocarbonols in an amount sufficient to improve the stability of the plating bath and the quality of the copper plate obtained from the bath but less than the amount which prevents deposition of copper from the bath.

2. The improvement of claim I wherein the bisulfite adduct is used in conjunction with a water-soluble borate.

3. The improvement of claim 1 wherein the water-soluble salt is the bisulfite adduct of an unsaturated lower aliphatic hydrocarbonol.

4. The improvement of claim I wherein the water-soluble salt is the bisulfite adduct of an acetylenic lower aliphatic hydrocarbonol.

5. The improvement of claim I wherein the water-soluble salt is the bisulfite adduct of l,4-butynediol.

6. The improvement of claim 1 wherein the water-soluble salt is the bisulfite adduct of l,4-butynediol used in conjunc tion with a water-soluble borate.

7. The process of electroless plating of copper which comprises contacting a substrate catalytically active in causing deposition of copper from electroless copper plating solutions at least on selected areas thereof with the plating composition of claim 1.

8. The process of claim 7 wherein after the deposition of copper on the substrate, the copper plated substrate is annealed to further increase the quality ofthe copper plate.

9. The process of electroless plating of copper which com prises contacting a substrate catalytically active in causing deposition of copper from electroless copper plating solutions at least on selected areas thereof with the plating composition of claim 6.

10. The process of claim 9 wherein after the deposition of copper on the substrate, the copper plated substrate is annealed to further increase the quality of the copper plate.

Claims (9)

1. 2. The improvement of claim 1 wherein the bisulfite adduct is used in conjunction with a water-soluble borate.
2. 3. The improvement of claim 1 wherein the water-soluble salt is the bisulfite adduct of an unsaturated lower aliphatic hydrocarbonol.
3. 4. The improvement of claim 1 wherein the water-soluble salt is the bisulfite adduct of an acetylenic lower aliphatic hydrocarbonol.
4. 5. The improvement of claim 1 wherein the water-soluble salt is the bisulfite adduct of 1,4-butynediol.
5. 6. The improvement of claim 1 wherein the water-soluble salt is the bisulfite adduct of 1,4-butynediol used in conjunction with a water-soluble borate.
6. 7. The process of electroless plating of copper which comprises contacting a substrate catalytically active in causing deposition of copper from electroless copper plating solutions at least on selected areas thereof with the plating composition of claim 1.
7. 8. The process of claim 7 wherein after the deposition of copper on the substrate, the copper plated substrate is annealed to further increase the quality of the copper plate.
8. 9. The process of electroless plating of copper which comprises contacting a substrate catalytically active in causing deposition of copper from electroless copper plating solutions at least on selected areas thereof with the plating composition of claim 6.
9. 10. The process of claim 9 wherein after the deposition of copper on the substrate, the copper plated substrate is annealed to further increase the quality of the copper plate.